make_ttsplit: make training/test splits

In kevinmhadi/khtools: Making Mskilab Life Easier

make training/test splits

Description

training and testing

Usage

make_ttsplit(
  dat,
  field = NULL,
  use.index = TRUE,
  k = 10,
  nested = FALSE,
  times = 1,
  transpose = TRUE,
  seed = 10,
  partition_prob = 0.632,
  split_type = c("crossfold", "resample", "partition"),
  as.data.table = FALSE,
  rngkind = "L'Ecuyer-CMRG"
)

Arguments

dat	data.table or data.frame of one row per observation. assumes each observation is a row
k	number of groups

Value

a list of row ids corresponding to each fold and training and test split

Author(s)

Kevin Hadi

kevinmhadi/khtools documentation built on Jan. 16, 2025, 4:18 p.m.